{"responseHeader":{"status":0,"QTime":3,"params":{"q":"{!q.op=AND}id:\"101834\"","hl":"true","hl.simple.post":"","hl.fragsize":"5000","fq":"!embargo_tdt:[NOW TO *]","hl.fl":"ocr_t","hl.method":"unified","wt":"json","hl.simple.pre":""}},"response":{"numFound":1,"start":0,"docs":[{"file_name_t":"Malley-A_Shading_Method.pdf","thumb_s":"/42/1c/421ca9d41b8da223586a6c73494929695029a903.jpg","oldid_t":"compsci 9947","setname_s":"ir_computersa","restricted_i":0,"format_t":"application/pdf","modified_tdt":"2016-05-25T00:00:00Z","file_s":"/b8/a5/b8a5315251123b67b498716df0d4cf57e4326862.pdf","title_t":"Page 50","ocr_t":"38 Since rays are traced out in scanline order, there is a coherency in the queries to the octree (and as a result, to the hash table). To take advantage of this coherency, items in the buckets of the hash table were reordered as the queries were made. The buckets were blocks of pointers to octree node structures. When an octree node pointer was 'hit' through the hash table, its pointer was moved up in the bucket so the next access of that node would be faster. 5.2 Octree Ray Intersection Complexity It is important to establish worst case bounds on ray intersections being performed with the octree method. An octree contains nt triangles, and the octree is constructed with a capacity of c triangles per node before subdivision of the node will take place. It is assumed that the triangles are uniformly distributed and that each of the triangles intersects only a single node. It is also assumed that the maximum depth for the octree dis such that d 2:: log8( ntf c) . As a result, the case where the depth restriction on the octree interferes with the node capacity restriction is infrequent. This allows consideration of an octree with a fixed maximum depth, d, and a worst case node capacity c. As implemented, the algorithm first classifies a ray starting point, an operation which is 0( d). All triangles in the corresponding node are tested, an operation which is 0( c). In the worst case, none of the c triangles tested in the node is intersected. The ray then extends into the next node of the octree, which involves another octree point classification. The worst case number of octree nodes visited occurs when the octree is full depth on all branches, which turns out to be a three-dimensional grid. This would provide gd leaf nodes of the octree. At most, by drawing a line through the grid, approximately gd/3) nodes of the octree will be visited. So, since the maximum number of objects in an octree node is assumed constant, the ray intersection time is the time to classify a point in the octree multiplied","id":101834,"created_tdt":"2016-05-25T00:00:00Z","parent_i":101866,"_version_":1679953745199235073}]},"highlighting":{"101834":{"ocr_t":[]}}}